Pancreatic ductal adenocarcinoma (PDA) is an unusually lethal disease with the highest 1-year and 5-year mortality rates of any cancer. An inability to detect the disease early together with multiple mechanisms of resistance to chemical and radiotherapies contribute to the extreme lethality of PDA. Insufficient tools exist for the early diagnosis, staging and treatment response of PDA. Current imaging modalities and invasive biopsies have not affected survival rates. The proposed study would break new ground by developing a non-invasive pancreatic assay for the isolation and analysis of pancreatic tumor cells from blood. Solid tumors are known to shed cells into the blood stream. The proposed strategy is a departure from traditional circulating tumor cells (CTCs) technologies in that the CTCs are purified from blood by labeling with a variety of antibodies expressed on PDA cells. The proposed strategy results in extremely high sensitivity and purity to enable molecular diagnostic assays on small numbers of cells. The current FDA-approved CTC enrichment technique collects cells expressing only relatively high levels of EpCam which is known to be highly variable and low in some tumor types. This proposal will use extensive knowledge of antigens expressed in PDA and test the specificity of their cognate antibodies on human blood, cancer cell lines and murine tumor model samples to develop a sophisticated antibody mixture for greater sensitivity in labeling pancreatic CTCs. Unwanted blood cells are differentially labeled and a microfluidic device is used to sort out the pancreatic CTCs for analysis. This proposal will also develop assays to analyze purified PDA CTCs for clinically relevant mutations in, for example, KRAS, TP53 and DPC4. The pancreatic blood biopsy will be further refined in well characterized genetically engineered mouse models of PDA that phenocopy defined genetic subtypes of the disease. The effort will be focused on maximizing the sensitivity and specificity of pancreatic CTC recovery and on building assays to track relevant mutations in PDA. The pancreatic blood biopsy will have immediate use as a discovery tool in the mouse model where key hypotheses on the disease initiation, progression and treatment are being made. More importantly the blood biopsy is designed for use in human population studies. Preliminary results with 10 ml of human donor blood demonstrate high levels of purity and compatibility with analysis of genetic mutations, cell signaling states and potentially other biomarkers of cancer progression and treatment response. It can be reasonably hoped that the final diagnostic tool would enable clinicians to make molecular diagnoses non-invasively and monitor for minimal residual disease, disease progression, and response or resistance to treatment.